Save ngraph.graph to dot format.
// Let's say we have a graph with two edges:
var graph = require('ngraph.graph')();
graph.addLink(1, 2);
graph.addLink(2, 3);
// Now save it to dot format:
var toDot = require('ngraph.todot');
var dotContent = toDot(graph);
This will set dotContent
to a string with a graph, described in dot format:
digraph G {
1 -> 2
2 -> 3
}
What to do with this? dot
is very well supported format by multiple graph
analysis software (e.g. gephi, graphviz).
You can easily transfer ngraph.graph
to your favorite platform!
To read it back to ngraph.graph
, use ngraph.fromdot
module:
var fromDot = require('ngraph.fromdot');
var newGraph = fromDot(dotContent);
Now newGraph
is an instance of ngraph.graph
The library will store attributes of nodes/edges, with a few limitations
- Attribute must be an object and not a primitive type
- Composite nested objects are JSON.stringified.
var graph = require('ngraph.graph')();
graph.addNode(1, { name: 'ngraph' });
graph.addLink(1, 2, { version: '42' });
// Now save it to dot format:
var toDot = require('ngraph.todot');
var dotContent = toDot(graph);
// you can parse it back:
var fromDot = require('ngraph.fromdot');
var restored = fromDot(dotContent);
// and expect attributes to be present:
restored.getNode(1).data.name === 'ngraph'
restored.getLink(1, 2).data.version === '42'
By default, when you toDot(graph)
the output is buffered into an array,
and flushed at the end. This may not be feasible for huge graphs, since it takes
extra memory.
For this uses case ngraph.todot
exposes a low level method write(graph, writer)
which
allows your code to own how actual output is stored/processed.
For example:
var graph = require('ngraph.graph')();
graph.addLink(1, 2);
graph.addLink(2, 3);
var toDot = require('ngraph.todot');
toDot.write(graph, function customWriter(line) {
console.log(line);
});
This will print dot file on the console without using extra memory:
digraph G {
1 -> 2
2 -> 3
}
With npm do:
npm install ngraph.todot
MIT